Hypokalaemia commonly occurs in acute myocardial infarction (AMI) and may be caused by elevated serum levels of adrenaline, allegedly by beta 2-adrenergic mediated influx of potassium (K) into cells. We investigated the effect on serum K of intravenous acebutolol (a relatively beta 1-selective agent) in 50 patients with AMI. Serum K was measured before and 1 hour after drug administration. The same measurements were made in a comparable control group of 30 patients who did not receive the drug. Mean serum K rose from 3.58 to 3.81 mEq/l (P less than 0.005) in the treated group. No significant change occurred in the control group. The rise in serum K could not be correlated with prior beta-blocker therapy, zone of infarction, prior diuretic therapy, or gender of the patient. We conclude that the administration of intravenous acebutolol after AMI raises serum K, despite the fact that this beta receptor blocking agent is relatively beta 1-selective. Since hypokalaemia is associated with an increased risk of ventricular fibrillation, it should no longer be assumed from acute intervention trials with beta-blockers in AMI which had mortality or arrhythmias as end-points, that beneficial effects were necessarily due to limitation of infarct size or to a direct anti-arrhythmic action of the drugs. Future trials should take the effect on serum K levels into consideration.